The Impact of COVID-19 Pandemic on

Small and Medium Enterprises

in Bangladesh

Asadul Islam
Director, Centre for Development Economics and Sustainability (CDES)
Professor, Department of Economics, Monash University

Atiya Rahman
Senior Research Associate, BRAC Institute of Governance and
Development (BIGD), BRAC University

Rafia Nisat
Research Associate, BRAC Institute of Governance and
Development (BIGD), BRAC University

December 2020
Development Economics Series 01

WORKING PAPER | 2021 | 1

2 | WORKING PAPER | 2021
Abstract
Like other economic players, the
novel pandemic severely hit small
businesses—the larger source of
growth and employment but also the
most vulnerable sector—by disrupting
national and international business
networks, supply chain, and demand.
To understand the evolving state of
small enterprises during pre, par,
and post-lockdown periods, BIGD in
collaboration with Monash University,
Australia conducted a survey on small
enterprises, mostly light-engineering
firms, and young workers across 18
districts in Bangladesh. The study
finds that lockdown measures caused
the majority of small enterprises shut
down, and during the early period
of relaxing the lockdown, one-third
of the enterprises were operating at
limited capacity. Demand drop and
the burden of fixed costs to run the
businesses were the prominent reasons
behind the drastic fall in profit. As a
result, workers were losing jobs and
the gender gap was widening, because
female labour-intensive work (i.e.
beauty parlour, tailoring) was affected
harder. Other findings of concern
include the emerging vulnerabilities for
the enterprises with lower endowment
and poor access to government
stimulus packages, and other financial
support. The study emphasises on
the importance of concrete targeting
criteria and support delivery platforms
to assist more vulnerable enterprises.
Finally, it highlights that the enterprises
that received BRAC’s intensive training
on occupational health and safety
(OHS), along with business training
and financial linkages, made double
profit compared to their counterparts.
However, the absolute amount of profit
was substantially lower for both groups,
compared to their pre-COVID profit,
indicating the importance of scaling up
such intensive training to create more
resilient enterprises in such crises.
WORKING PAPER | 2021 | 3
 Contents
Abstract

Acknowledgment

1. Introduction and Background 1

2. The Survey 3

3 Results from the Enterprise Owners’ Survey 4

3.1 Respondents’ profile 4
3.2 Business operations, sales and expenses 4
3.3 Consequences for hired workers 9
3.4 Challenges faced during the lockdown 9
3.5 Coping 10
3.6 Access to stimulus package by government 11
3.7 The time required for business recovery 12
3.8 Health guidelines and COVID-19 symptoms 13
3.9 How the BRAC intervention recipients are performing
compared to their counterparts during this pandemic 13

4. Results from the Workers’ Survey 16

4.1 Workers’ employment status and income relative to

the pre-COVID level 16
4.2 How the intervention recipient and non-recipient
workers are doing during this pandemic 18

5. Conclusion 19

References 20

Annexe 21

4 | WORKING PAPER | 2021

Acknowledgment
We would like to express our deepest gratitude to Dr. Narayan Das, Senior Research
Fellow, BRAC Institute of Governance and Development (BIGD); Imran Matin,
Executive Director, BIGD; Margaret Triyana, and Xing Xia for their continued support,
encouragement and valuable suggestions at different stages of the study. We are
also indebted to the survey respondents for sharing their time and useful information
for the study during this unprecedented time, without which this report could not be
generated. The field and data management team of BIGD also deserve special thanks
for their strenuous work. We thank Nusrat Jahan, Head of Business Development
and Knowledge Management for helping to prepare the report. We acknowledge the
generous financial support of GLM|LIC program supported by UK Aid.

WORKING PAPER | 2021 | 5

6 | WORKING PAPER | 2021
1. Introduction and Background
The COVID-19 pandemic has had
dramatic ramifications around the world.
To limit the spread of the disease, many
countries adopted lockdown and social
distancing measures. Although vital in
containing the virus, these measures
have also precipitated an unprecedented
economic crisis. At present, many
countries have started relaxing the
lockdown to restart economic activities,
which is expected to remain subdued for
some time. Bangladesh started easing
lockdown measures in early June,
but the movement of people across
the country and economic activity is
expected to remain restricted, as the
number of COVID-19 cases have not
yet decreased. Lockdown and social
distancing measures have hit Small
and Micro Enterprises (SMEs) especially
hard. These SMEs account for a large
portion of production and employment
in developing countries.
This study examines the impact of the
pandemic on SMEs and their workers
in Bangladesh. It takes advantage of
an ongoing project, implemented by
BRAC, with funding support from the
European Commission. The research
team previously collaborated with BRAC
to design and evaluate an intervention
in light engineering (LE) firms, using a
large scale randomised controlled trial
(RCT) that started in 2017. Most firms
in the LE sector are small and informal,
with high dependence on low-skilled
labour and little occupational health and
safety (OHS) measures. Most of their
production process involves welding or
working with hazardous substances,
which can result in frequent work-related
injuries and fatalities. The sector acts
as a support industry to other industries
at various stages of the supply chain.
The LE sector is one of the largest sub-
sectors of SMEs with two million workers,
contributing to 2% of GDP.
In our initial project, we analyse the
impact of intensive decent-work-
environment training on 2248 light
engineering (LE) firms. The RCT consists
of the following treatment arms: T1:
Managers/owners of firms receiving
intensive training on OHS; T2: OHS +
business training and financial linkages;
and C: firms in the control group with
no training. For the evaluation, we
conducted a firm survey in November-
December, 2019.
In this study, we seek to understand the
current situation of these SMEs and their
workers, and compare it with the pre-
COVID period. We examine whether
the effects of the lockdown measures
vary by firms’ decent-work-environment
training status. We examine whether
and how decent-work-environment
training helped to maintain a healthy
work environment for workers during
the pandemic. We hypothesise that
the treated firms are more likely to
adopt safety measures to minimise the
transmission of COVID-19. The first
round of the survey has been completed.
In the follow-up surveys, we will
examine the longer-term impact of the
COVID-19 pandemic on firm survival,
growth, investment, and profitability.
WORKING PAPER | 2021 | 1
 We will also examine the longer-term
effects of the pandemic on workers’
wellbeing and labour market outcomes.
We will compare these outcomes with
information already collected in 2019,
before the pandemic.
The rich literature on microenterprise
development in LMICs has largely
focused on either credit or formal
business training, while we incorporate
OHS to promote worker wellbeing. Our
findings will contribute to the literature
on the broader effects of training for
firms and workers in the informal sector,
(Valdivia 2011, DeMel et al. 2014,
Karlan et al 2015, Fiala 2013, McKenzie
and Woodruff 2014, Brooks et al
2016) and the importance of providing
a decent work environment in LMICs
2 | WORKING PAPER | 2021
(Fields 2003). The interaction between
our intervention and the pandemic will
allow us to examine the role of OHS
and business training on firms’ ability to
survive and protect their workers.
In addition to these light engineering
firms, our sample also includes a
small sample of other enterprises such
as hotels, beauty parlours, clothing/
tailoring, and general stores. This report
documents the condition of the surveyed
small and medium enterprises (hence
after SMEs) at three stages—pre, par,
and post lockdown. We also aim to
explore the condition of the firms that
received BRAC intervention, (hence after
treatment) compared to those which did
not receive any intervention (hence after
control).
2. The Survey

We (re)surveyed all 2238 LE firms in In addition to measuring these

our RCT and 126 other enterprises1
provided by BRAC Bank. The survey
was conducted by BRAC Institute of
Governance and Development (BIGD),
Brac University. Despite the withdrawal
of the lockdown, there is still limited
mobility in Bangladesh. Hence, we
contacted survey respondents over the
phone. The length of each interview in
both rounds was kept to a minimum
to ensure data quality and reliability.
Each interview took 20-25 minutes. We
plan to conduct more waves of surveys
to understand the situation at different
intervals.
The survey covers four broad areas: (1)
enterprises’ economic behaviour and
economic outcome during and after
the lockdown (e.g. enterprise revenue,
output, number of days in production,
number of days shut down, and the
number of workers retained during);
(2) physical wellbeing, particularly
COVID-19 symptoms carried by the
workers and their family members; (3)
health and safety measures taken by the
enterprise or enterprise manager that
aims at minimising the risk of COVID-19
transmission, and (4) accessibility to
government initiatives.
important outcomes, we also aim to
study how each of these differ by the
initial treatment status. We compare
LE firms that received decent-work-
environment training to those that did
not, and answer questions such as: Do
treated firms fare better economically
during and after the pandemic? Are
treated firms more likely to adopt safety
measures that minimise the transmission
of COVID-19? Do workers in treated
firms experience better physical
wellbeing during the pandemic?
Besides this survey, we also interviewed
the apprentices who received training
from BRAC on the LE services and their
non-recipient/control counterparts. As a
part of the RCT mentioned earlier, these
apprentices were previously surveyed
twice. We treat them as workers in this
report.

1
There are few characteristics of these firms before the
pandemic. About 12% of these enterprises are run by
women. The average age of these owners is about
38 years. These enterprises were comprised of six
workers and on average their profit was 56 thousand
before the pandemic.

WORKING PAPER | 2021 | 3

 3. Results from the Enterprise
Owners’ Survey

3.1 Respondents’ Profile Finally, the remaining samples represent

hotels (1%), beauty parlours (1%),
We successfully interviewed 19602 clothing and tailoring stores (2%), and
out of 2364 SMEs from 14th to 23rd July, general or variety stores (1%).
2020. Almost all respondents were male
(99.13%) with an average age of 42
years. Only 4% of them moved to their Figure 1. Distribution of samples by
current residences after lockdown. Our enterprise type
surveyed enterprises are widely spread
out across 18 districts of Bangladesh,
mostly Bogra (13%), Gazipur (8.4%),
Jessore (7.9%), Mymensingh (6.2%), and
Cumilla (6.1%).

Different SME owners were the

respondents of this survey. As mentioned
earlier, most of our surveyed samples
(95%) belong to the LE sector, and
we have three-round panel survey
information for the pre-COVID (2017,
2018, and 2019) period for these firms.
About one-third of the respondents
(32%) run automobile spares businesses, 3.2 Business Operations, Sales
while 25% and 16% were grills and and Expenses
agriculture machines, accessories, and
spares firms, respectively. 22% of LE Sixty-nine percent of the surveyed
enterprises were jute and textile machine enterprises were closed during two
and spares; engineering and metal months of lockdown (March 26-May 31,
industry machines and spares; lathe 2020). Only 2% were fully operational
firms; etc. and 29% were partially open. After
the economy reopened, the scenario
reversed. 61% of the enterprises were
2
In total, 1990 respondents gave consent to
fully open and 37% were partially open.
participate in the survey, of which, 30 respondents Only 2% were closed after the lockdown
did not run any enterprise. Thus, the survey did not lifted.
record any information on their business before,
during and after lockdown period.

4 | WORKING PAPER | 2021

Figure 2. Business Operation Status
Before lockdown, these firms used
to operate 26 days a month, with an
average of 11 working hours a day.
During lockdown, these enterprises
operated only 14 days a month, with
an average of four working hours a
day. After the end of the lockdown, the
Figure 3. Working days and hours
working day increased to 22 and the
working hours increased to eight. These
figures are still lower than pre-lockdown,
because slightly more than one-third
(37%) of the enterprises were not fully
operational after the lockdown was
lifted.
WORKING PAPER | 2021 | 5
 Looking into the operation status by the
types of enterprises, we find that the
distinctive service sector enterprises i.e.
beauty parlours, tailoring, etc., where
maintaining social distance is quite
difficult, were almost completely shut
down during the lockdown and were less
likely to reopen afterwards.
Now, the question is what the sales
status and expenses are. Compared
to February, (before the lockdown
period) sales dropped by 76% during
lockdown and 52% in July (after the end
of lockdown). The drop in expenses are
lower than that in sales, both during and
after lockdown. As a consequence, we
find that the enterprises are running at
break-even point even after the economy
opened up.

Figure 4. Business Operation Status by Enterprise Type

Figure 5. Sales, Operating Cost and Total Profit

6 | WORKING PAPER | 2021

Looking into the drop in sales by
the enterprise type, we find that the
distinctive services sectors such as
clothing/tailoring shops, and beauty
parlours are facing a severe decline in
their sales, indicating a lower demand
for non-essential services. Moreover,
there might be two-fold drivers. First,
people are aware that it would be
difficult to maintain physical distance
while receiving these services. Second,
another BIGD study (Rahman et al.
2020) shows that a significant number
of vulnerable non-poor households—
who represent a large body of customers
of the service sectors—fell below the
poverty line during the lockdown and
Figure 6. Drop in Sales by Enterprise Type relative to pre-COVID
Figure 7. Rent and Utility Bill Payment Status
remain there even after the economy
reopened.
After the lockdown was lifted, we find
that the owners are reopening their
enterprises. They are, however, facing
difficulties paying rent and utility bills.
22% and 13% of the owners were
unable to pay rent and utility bills,
respectively. This may be due to two
reasons. First, the owners are now
paying for variable costs to run their
businesses. Second, they may be able
to negotiate with their landlords and
pay rent later, since they are back in
business.
WORKING PAPER | 2021 | 7
 Exploring which enterprises are more
likely to reopen after the end of
lockdown, we find that those which had
higher initial capital in 2017 are more
likely to operate fully at present.
Moreover, the relatively rich enterprise
owners faced a lower drop in sales,
compared to their pre-COVID levels.
While the poor enterprises experienced
a 49% drop, the rate is 25% among the
richest enterprises. Such figures point
out that the poorer entrepreneurs are at
higher risk.

Figure 8. Operating Status by initial capital categories

Figure 9. Drop in sales relative to pre-COVID by initial capital categories

8 | WORKING PAPER | 2021

3.3 Consequences for hired
workers
About 24% of business owners laid off
their workers during the lockdown, while
the rate has reduced to 11% after the
end of lockdown. Interestingly, almost all
enterprises (98%) had their employees
work and get paid for fewer hours
during lockdown, and the current rate
(81%) is also quite high in this regard.
This is similar to the finding we showed
earlier, that a significant portion of the
firms are partially reopening. They are
operating for fewer hours, resulting in
workers facing a loss of income.
Figure 10. Consequences for hired workers
Figure 11. Disruption during the lockdown
3.4 Challenges faced during
the lockdown
Three-quarters of the owners reported
that they faced disruption or extreme
disruption in terms of receiving orders.
39% of the surveyed owners said that
they had to shut down operations
due to temporary lockdown. Similarly,
40% noted that they were unable to
pay employees and maintain business
operations.
WORKING PAPER | 2021 | 9
 Correspondingly, 76% reported having
no income, due to receiving no orders,
as one of the main problems they faced
during the lockdown period, and 50%
reported paying salaries to employees as
one of the main issues.
3.5 Coping
of raw materials. Subsequently, 25%
of the owners reported that they are
reducing production to cope with the
shortage. Other prominent strategies are
increasing product prices (17%), seeking
new procurement channels (14%), and
delaying goods delivery (13%).

Sixty-one percent of the surveyed

owners reported having a shortage

Figure 12. Major problems faced during the lockdown

Figure 13. Coping mechanisms for the shortage of raw materials

10 | WORKING PAPER | 2021

Slightly less than half of the owners
(47%) have access to business loans or
grants to support business recovery and
66% of the surveyed owners are taking
measures or planning to respond to
business disruption due to the pandemic.
Looking into loan access by the
enterprise type, we find that the
distinctive service sectors are at greater
risk with the lowest access to a loan for
business recovery (23%).
We look into the combination of access
to loans and recovery plans and
find that one-fourth of our surveyed
enterprises do not have access to loans
and do not have any plan to recover
their businesses (Figure A1 in Annexe).
This is alarming and highlights the
importance of introducing business
training for enterprises.
3.6 Access to stimulus
package by Government
On 13th April, the Bangladesh
Government announced stimulus
Figure 14. Access to loans/grants to support business recovery by
enterprise type
packages to support SMEs. Three
months after this declaration, only 63%
of the surveyed workers knew about the
stimulus package.
Eighty-one percent of those who know
about this package reported that it is
quite or very difficult to get the support
and 17% reported that they know how
to get the support. Interestingly, 31%
of those who know, noted that they
have already applied for the package.
Unfortunately, only one owner received
it. There arises the question of efficiency.
Data shows that only 65 out of 1960
enterprises have applied for support.
Exploring the differences between the
firms which have and have not applied,
we find that resilient ones are more
likely to do so. More than a third of the
enterprises that have applied for this
package were open during lockdown,
while 30% were not. Moreover, the
amount of current sales of those who
have applied is about 4% higher than
that of those who have not.
WORKING PAPER | 2021 | 11
 During the pandemic, government
support is of utmost importance.
Like other formal financial institutes,
microfinance institutes are also
experiencing a distressing period. Thus,
the Government should reach out to
these small and micro-entrepreneurs
in a more efficient manner. Fewer
formalities to receive the Government
support package might make the path
easier for entrepreneurs.
3.7 The time required for
business recovery
About three-fourths of the owners are
extremely worried about the future of
their business. Moreover, 45% of the
owners are uncertain about when they
will be able to operate their business at
full capacity.

63% know about

the Govt. Initiative

81% reported that it is

17% know how to
quite/very difficult to
get the support
get the support

31% applied for 1 owner received

the support the support

Figure 15. Government Support

Figure 16. Time required to fully recover

12 | WORKING PAPER | 2021

3.8 Health guidelines and
COVID-19 symptoms
Owners’ reported data reveals that
the workers of the distinctive service
sectors, where it is difficult to maintain
social distance, are more likely to have
COVID-19 symptoms i.e. fever, cough,
shortness of breath, or muscle pain.
Furthermore, we explore whether
businesses are being closed because
their workers are running a higher risk
of getting infected. We find that there is
no correlation between the businesses
opening and the workers showing any
COVID-19 symptoms.
3.9 How the BRAC
intervention recipients are
performing compared to
their counterparts during
this pandemic
Before comparing the pre, par, and post-
lockdown situations of treatment and
control groups, we aim to understand
Figure 17. COVID-19 symptoms
the characteristics of the attrited firms
during the current phone interview. We
could not reach 395 out of 2238 LE
firms to interview over the phone. Of the
attrited firms, 43% could not be reached,
as their provided contact numbers were
switched off, while 24% did not provide
their consent to be interviewed (Figure
A3). Moreover, 41% of these firms
belonged to the treatment group and
the rest were in the control group. As
mentioned earlier, we surveyed these LE
firms in December 2019-January 2020
as a part of the evaluation design using
an RCT. Utilising that dataset, we intend
to understand what type of firms were
attrited during the current survey. Out
of these 395 attrited firms, 252 were
successfully interviewed in December
2019-January 2020. While analysing
their characteristics at that time, we find
that 64% of them were engineering
and metal industry machine, a1919nd
spares workshops (Figure A4).
Furthermore, their initial capital was BDT
245 thousand and the amount of sales
was BDT 103 thousand at the end of
2019/early 2020.
WORKING PAPER | 2021 | 13
 Table 1 delineates the differences
between treatment and control groups
during pre, par, and post-lockdown.
While comparing the treatment firms
with their control counterparts, we did
not find significant differences between
these groups, in terms of the relevant
indicator (Table 1). The difference
between treatment and control firms,
in terms of the pre-COVID profit level
is 14.7% of the control group’s mean,
with a higher level for the treatment
group. In addition, the profit level of
treatment firms is higher during and
after the lockdown period (Figure 18).
However, the absolute amount of profit
for the treatment firms was lower post-
lockdown, compared to during pre-
lockdown. Given the higher profit levels
of the treatment firms compared to their
counterparts across all periods, it is also
logical that they are more capable of
hiring workers. Consequently, Figure A5
suggests that the number of employees
working in the treatment firms is indeed
higher, even during lockdown. We also
find that compared to the treatment
firms, control firms are working more
days and longer hours (Figure A6 and
A7).
Although the percentage of control firms
who know about the Government’s
recently declared stimulus package is
slightly higher, more treatment firms
have applied for support (Figure A8).
Furthermore, as reported by the owners,
the workers of the treatment firms are
less likely to carry COVID-19 symptoms.
The training on OHS might play a role to
enhance their awareness (Figure A9).

Table 1. Intervention recipient vs. non-recipient enterprises

Indicators Treatment Control Difference

Profit before lockdown (BDT) 30003.29 26150.69 3852.6 (2216.32)

Profit during lockdown (BDT) 1333.5 -163.403 1496.90 (1351.49)

Profit after lockdown (BDT) 2456.37 1916.23 540.13 (4089.5)

Know about government initiative (%) 62.24 63.66 0.0142 (0.0225)

At least one worker has COVID-19 symptoms 23.67 26.92 0.0233 (1.447)
(% of Workshops)

Note: Standard errors in parentheses. We have used t-test to test the differences.

14 | WORKING PAPER | 2021

 40000
Average value of profit in the past one month
0 10000 20000 30000

Before lockdown During lockdown After lockdown

Time period
treatment control

Figure18 . Average value of profit

To assess whether BRAC’s intervention coefficient for the intervention parameter

program has any significant effects is positive for profit, both during and
on the profit levels during and after after lockdown, but significant only for
lockdown, we run an ordinary least profit during the lockdown period. This
squares regression controlling for basic suggests that the intervention aided the
firm characteristics and profit, during participants in coping better during such
the follow-up survey conducted in the a difficult time.
December-January period. We also
cluster the errors at the market level.
The results are given in Table A1. The

WORKING PAPER | 2021 | 15

 4. Results
from the Workers’ Survey
Out of 1652 samples, we
successfully interviewed 1014
youths. Out of 1014, 798
(79%) were employed before
the pandemic. This report
focuses on these 798 workers.
Female, 21.80%
Male, 78.20%
Figure 19. Distribution
of workers employed in
February
Twenty-two percent of the
employed workers in February
were female. The average age
of these surveyed workers is 23
years.
16 | WORKING PAPER | 2021
4.1 Workers’ employment status and
income relative to the pre-COVID level
Fifty-eight percent of the employed males
were hired during the lockdown, while the rate
increased to 85% post-lockdown. As shown in
figure 19, female workers were more likely to lose
their jobs during the lockdown and less likely to
be rehired afterwards. As shown earlier, female
labour intensive enterprises, i.e. clothing, tailoring,
and beauty parlours experienced a larger drop
in their sales, even after the economy opened
up (figure 6). As a result, female labour force
participation is also being hampered.
Figure 20. Workers’ employment status at par
and post-lockdown
Compared to the pre-COVID level, income
dropped 60% and 65% for male and female
workers, respectively. Income recovery for male
workers is higher, compared to their female
counterparts. Female workers are facing a 52%
drop in income after the end of the lockdown and
their income recovery rate is very poor.
There is always a gap between the
wages of male and female workers.
Male workers get higher pay compared
to females. During this pandemic, the
gap is widening. Female workers were
working more hours during and after
lockdown to survive, but, the earning per
day gap was higher, compared to the
pre-COVID status (Figure 21).
Slightly more than 80% of current
workers reported that they regularly
wear masks on the job, and 70%
maintain social distance. However, there
are no significant differences between
Figure 21. Drop in workers’ income
Figure 22. Per day earning gap between male and female workers
male and female workers in terms of
these compliances.
Interestingly, about 24% of the workers
noted that no one at their workplaces
had COVID-19 symptoms in the last
three months. This rate is quite close
to the percentage reported by the
enterprise owners (26%).
Almost 40% of the workers reported that
they might get infected by COVID-19,
but they cannot stop working for a living
(Figure A2 in annexe).
WORKING PAPER | 2021 | 17
 Figure 23. Maintaining health guidelines at the workplace

4.2 How the intervention recipient and non-recipient workers are doing
during this pandemic

As mentioned earlier, these workers lockdown higher among the intervention

received intensive decent-work-
environment training from BRAC. Due to
this training, the intervention recipients
(i.e. treatment) might be less likely to
work during the lockdown. However,
after the lockdown lifted, about 80%
of them returned to work, making
the employment recovery rate after
recipients. In terms of compliance with
COVID-19 related health guidelines, i.e.
wearing masks, gloves, and maintaining
social distance, both intervention
recipients and non-recipients reported
similar rates.

Table 2. Intervention recipient vs. non-recipient workers

Indicator Treatment Control Difference

Employed in April (% of employed workers in Feb) -9.549***
52.49 62.03
(3.629)
Employed in July (% of employed workers in Feb) 80.32 78.64 1.674 (2.953)
Monthly income in April (BDT) 2827.58 3392.10 -564.5 (352.9)
Monthly income in July (BDT) 5404.58 4986.43 418.1 (456.4)
Regularly wear a mask during work (% of current workers) 84.55 81.22 3.335 (3.287)
Regularly wear gloves during work (% of current workers) 14.52***
36.86 22.34
(4.071)
Regularly maintain 1.5 meters distance from co-workers -0.159
during work (% of current workers) 71.75 71.91
(4.142)

Note: Standard errors in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01. We have used t-test to test the
differences.

18 | WORKING PAPER | 2021

5. Conclusion
Our study attempts to assess the
conditions of SMEs in Bangladesh in
the context of the ongoing COVID-19
pandemic. As our sample is from an
ongoing project implemented by BRAC,
where we collaborated to analyse the
impact of their decent-work-environment
training on light engineering (LE) firms,
we can make further comparisons
between enterprises who received the
intervention and those who did not.
Results suggest that these small and
informal enterprises are encountering
massive drops in their sales, which is
about 55% relative to the pre-COVID
level. Given the relatively lower drop
in expenses, the findings imply that the
firms are operating at their break-even
point despite the economy being open.
Quite alarmingly, we find that relatively
poorer enterprises, i.e. those with lower
initial capital, are less likely to operate
at full capacity after the reopening of
the economy and more likely to face
a severe drop in sales. Furthermore,
information on government stimulus
package does not reach the SMEs
adequately. In this case, enterprises with
a higher profit before the pandemic are
more likely to apply for such support.
Such figures raise the risk of relatively
smaller firms being left behind.
Finally, by comparing enterprises who
received the intervention from BRAC
to their non-recipient counterparts, we
find that the profit level of the treatment
enterprises is higher than that of the
control enterprises, par and post-
lockdown, despite being more likely
to remain closed or operate partially
during lockdown. COVID-19 symptoms
of workers in treatment enterprises
are also lower. This insight from our
descriptive analysis indicates that
enterprises who received training are
better able to cope with the lockdown
and the pandemic as a whole.
Moreover, from analysing our data on
the workers, we see that the female
workers are at a higher risk.
Our study has some methodological
limitations including reporting biases
during these unprecedented times,
seasonality, and lack of heterogeneity
in our sample. As our data puts more
focus on the light engineering sector, we
cannot concretely say that our results are
representative of the entire SME sector.
We plan to conduct two more rounds
of surveys to explore the coping and
recovery dynamics of these firms.
The informal sector is one of the
major drivers of our economy and
small and medium enterprises are of
immense significance, in the long run,
to sustain our growth momentum and
reduce inequality. Thus, we require
comprehensive and inclusive strategies
to aid such firms in post-pandemic
business recovery and ensure that none
of firms fall behind.
WORKING PAPER | 2021 | 19
 References
Brooks, W., Donovan, K.M. and
Johnson, T., 2016. The Dynamics of
Inter-Firm Skill Transmission among
Kenyan Microenterprises. Helen
Kellogg Institute for International
Studies.
De Mel, S., McKenzie, D. and Woodruff,
C., 2014. Business training and
female enterprise start-up, growth,
and dynamics: Experimental
evidence from Sri Lanka. Journal
of Development Economics, 106,
pp.199-210.
Fields, G.S., 2003. Decent work and
development policies. International
Labour Review, 142(2), pp.239-262.
Fiala, N. (2013). Stimulating
microenterprise growth: Results
from a loans, grants and training
experiment in Uganda. Working
Paper.
tailors in Ghana. Journal of Economic
Behavior & Organization, 118:281–
302.
McKenzie, D. and Woodruff, C., 2014.
What are we learning from business
training and entrepreneurship
evaluations around the developing
world? World Bank Research
Observer, 29(1), pp.48-82.
Valdivia, M., 2011. Training or technical
assistance? A field experiment
to learn what works to increase
managerial capital for female
microentrepeneurs. Working Paper.
Rahman et al. (2020) https://bigd.bracu.
ac.bd/wp-content/uploads/2020/08/
PPRC-BIGD-Rapid-Response-Research-
Phase-II-Findings.pdf.

Hasle, P., & Antonsson, A.-B. (2009).

Informal Sector and Small
Enterprises. In K. Elgstrand, &
N. Petersson (Eds.), OSH for
Development (pp. 587-604).
Stockholm, Sweden: Royal Istitute of
Technology.

Karlan, D., Knight, R., and Udry, C.,

2015. Consulting and capital
experiments with microenterprise

20 | WORKING PAPER | 2021

Annexe

Figure A1. Access to loan and recovery plan

Figure A2. Workers’ perception of being infected by COVID-19

WORKING PAPER | 2021 | 21

 Figure A3. Reasons Behind Attrition

Figure A4. Business type of attrited firms according to follow up survey

(% of firms)

22 | WORKING PAPER | 2021

 Average no. of workers in the past one month
1.5 2 2.5 3

Before lockdown During lockdown After lockdown

Time period
treatment control

Figure A5. Average no. of workers (Treatment vs. control)

Average no. of days worked in the past one month
10 15 20 25

Before lockdown During lockdown After lockdown

Time period
treatment control

Figure A6. Average no. days worked in a month (Treatment vs. control)

WORKING PAPER | 2021 | 23

 12
Average no. of hours worked in a day in the past one month
10
8
6
4

Before lockdown During lockdown After lockdown

Time period
treatment control

Figure A7. Average no. of hours worked in a day (Treatment vs. control)
Percentage of firms who received government support
2 3 4 5

treatment control
Whether received intervention
95% confidence intervals

Figure A8. Percentage of firms applying for government support

(Treatment vs. control)

24 | WORKING PAPER | 2021

 Percentage of firms with at least one person showing COVID-19 symptoms
20 22 24 26 28 30

treatment control
Whether progress beneficiary
95% confidence intervals

Figure A9. Percentage of firms with at least one worker showing COVID-19
symptoms (Treatment vs. control)

WORKING PAPER | 2021 | 25

 Table A1. OLS regression results to assess BRAC intervention
impact on profit

OLS coefficients
Profit after lockdown Profit during lockdown
Whether PROGRESS beneficiary 0.19 (0.17) 0.23* (0.10)
Gender -0.02 (0.64) 3.62*** (0.22)
No. of workers in the last one month -0.00 (0.00) -0.11** (0.03)
Profit during December-January 0.09 (0.01)
***
-0.02 (0.05)
Initial capital according to follow up survey -0.01 (0.00)
*
-0.00 (0.00)
Agriculture, machines, accessories and spares 0.04 (0.11) 0.15* (0.06)
Grill business -0.03 (0.19) 0.52** (0.14)
Service sectors 2.09*** (0.39) 0.00 (.)
Other enterprises 0.16 (0.12) 0.06 (0.25)
Feni -1.03 (0.20)
***
-1.00** (0.34)
Brahmanbaria -0.39 (0.74) -0.26* (0.11)
Noakhali -1.70** (0.47) -0.76** (0.21)
Bogra -0.91** (0.21) -0.97** (0.28)
Jessore -1.36* (0.53) -0.94* (0.40)
Shatkhira -1.74** (0.45) -1.69** (0.41)
Barisal -0.18 (0.24) -0.82 (0.64)
Moulavibazar -0.66** (0.18) -0.35* (0.16)
Habiganj -1.82** (0.61) 0.80** (0.20)
Gazipur -2.66*** (0.51) -1.23*** (0.22)
Narayanganj 0.02 (0.26) -0.94 (0.46)
Kishoreganj -2.77** (0.88) -1.70*** (0.26)
Faridpur -0.82*** (0.15) -1.61*** (0.06)
Nilphamari -0.81 (0.38) -0.91*** (0.13)
Gaibandha -0.65 (0.59) -1.14* (0.46)
Rangpur -0.65* (0.27) -0.9 (0.55)
Mymensingh -0.32** (0.11) -0.36 (0.30)
Village/local shops -0.14 (0.10) -0.43** (0.11)
Sub district market -0.28 (0.18) -0.21 (0.11)
District level market -0.56*** (0.12) -0.34 (0.18)
Divisional market -0.66** (0.16) -0.45 (0.40)
Constant 1.65* (0.70) -2.34*** (0.15)
Observations 1665 532
Adjusted R 2
0.074 0.090

26 | WORKING PAPER | 2021